bug#18892: few test failure with 'grep-2.20.72-d512'

[Assaf Gordon]

On 10/30/2014 03:07 PM, Jim Meyering wrote:
> For the record, you may want to run something like this to survey
> whether \w matches E-acute across all  installed zh_CN locales:
>
> (here's output from an os x 10.9.5 system -- on debian unstable and
> fedora rawhide, they're all "0"s)
> $ for i in $(locale -a|grep -i cn); do printf '\303\251'|LC_ALL=$i
> src/grep -q '\w'; echo $?: $i;done
> 0: zh_CN
> 1: zh_CN.eucCN
> 1: zh_CN.GB18030
> 1: zh_CN.GB2312
> 1: zh_CN.GBK
> 0: zh_CN.UTF-8
===
Debian 7.6
(no output, installed locales are: C  C.UTF-8  en_US.utf8 POSIX)
====
Ubuntu 14:
(no output, installed locales are C C.UTF-8 POSIX and en_* , en_*.utf8)
====
FreeBSD 10.0
1: zh_CN.GB18030
1: zh_CN.GB2312
1: zh_CN.GBK
0: zh_CN.UTF-8
1: zh_CN.eucCN
====
FreeBSD 9.3
1: zh_CN.GB18030
1: zh_CN.GB2312
1: zh_CN.GBK
0: zh_CN.UTF-8
1: zh_CN.eucCN
===
OpenBSD 5.5
0: zh_CN.UTF-8
===
NetBSD  6.1.4
1: zh_CN
1: zh_CN.GB18030
1: zh_CN.GB2312
0: zh_CN.UTF-8
1: zh_CN.eucCN
===
CentOS 7
0: bo_CN
0: bo_CN.utf8
0: ug_CN
0: ug_CN.utf8
0: zh_CN
0: zh_CN.gb18030
0: zh_CN.gb2312
0: zh_CN.gbk
0: zh_CN.utf8
====
CentOS 6.5
0: bo_CN
0: bo_CN.utf8
0: ug_CN
0: ug_CN.utf8
0: zh_CN
0: zh_CN.gb18030
0: zh_CN.gb2312
0: zh_CN.gbk
0: zh_CN.utf8
=====
OpenSUSE 13.1
0: bo_CN
0: ug_CN
0: zh_CN
0: zh_CN.gb18030
0: zh_CN.gbk
0: zh_CN.utf8
=====